B2B SaaS

Which marketing numbers actually deserve a dashboard?

Written by
Pravin Kumar
Published on
Sep 24, 2026

Which marketing numbers actually deserve a dashboard?

Only numbers that would change a decision this month. For most small B2B teams that is five or six. Qualified conversations created, where they came from, conversion from conversation to opportunity, pipeline value, and one leading indicator per active channel. Everything else belongs in a report you open when you have a question.

Dashboards fail by inclusion. Every number someone once asked about gets added, nobody removes anything, and eventually the dashboard becomes wallpaper that everyone has stopped reading. A dashboard with thirty tiles is a dashboard with zero.

The discipline is deciding what earns permanent space, and being honest that permanent space is expensive attention.

What is a dashboard actually for?

It answers one question repeatedly. Is the thing we are doing working, and if not, where is it breaking. A dashboard is a monitoring tool, not an analysis tool, and the difference decides what belongs on it.

Monitoring means you look at the same numbers on the same cadence and notice change. Analysis means you have a specific question and you go digging. Those need different surfaces. When people mix them, the dashboard fills up with the residue of old questions.

So the test for a tile is simple. If this number moved by twenty percent, would anyone do anything differently? If the answer is no, it does not belong on the dashboard, no matter how interesting it is.

Which numbers make the cut for a small B2B team?

Start with the outcome and work backwards. Qualified conversations created this month, their source, the rate at which they become real opportunities, pipeline value created, and one early signal per channel you are actively working. That is five tiles and it tells you almost everything.

Qualified conversations is the number I would keep if I could only have one. Not form fills, not sign ups, not traffic. Conversations with someone who could plausibly buy. It is the first point where marketing effort becomes a business fact, and it is specific enough to argue about.

Source belongs next to it because the same total means different things depending on where it came from. Twenty conversations from referrals and twenty from a content channel demand different next moves, and averaging them hides the only useful information.

The leading indicator per channel is the part teams skip. For content, something like impressions on the pages you are actually working. For outbound, positive reply rate. For events, meetings booked on site. One per channel, chosen because it moves before the outcome does.

Which popular metrics do not deserve the space?

Total sessions, bounce rate, email open rate, follower counts, and most engagement composites. They move for reasons you did not cause, and they rarely change a decision. Keeping them on a dashboard trains everyone to explain noise.

Total sessions is the most seductive. It always has a story, it goes up when you publish, and it can be entirely disconnected from revenue. A month where sessions fall and qualified conversations rise is a good month, and a dashboard that leads with sessions will make that look like a problem.

Open rate has become genuinely unreliable as a signal of human attention, for reasons that have to do with how mail clients fetch images rather than with your copy. Treat it as a weak diagnostic at best, and never as a goal, because optimising toward it changes subject lines in the wrong direction.

I would also retire any composite score a tool invents. If a number cannot be explained in one sentence to a founder, it cannot be defended in a meeting, and it will eventually be used to argue for something nobody understands.

How do you handle attribution honestly on a dashboard?

Show it as a claim rather than a fact. Label the model you are using, keep one simple view rather than three competing ones, and record self reported source alongside the tracked one. Where a buyer says they heard about you is often better information than any tracking chain.

Multi touch reality does not fit in a tile. A prospect reads two articles, hears you on a podcast, searches your name, and fills in a form. Any single attribution model will credit one of those and hide the rest, so the useful move is to stop pretending precision exists and start looking at pattern.

My habit is to keep tracked source and self reported source side by side and treat disagreement as information rather than as an error to fix. If they disagree constantly, your tracking is missing a real channel, and that is worth knowing. A small team can get most of the way there with clean campaign tagging, which is why UTM discipline matters more than any dashboard design, and why attribution for a small B2B team should stay deliberately simple.

How often should each number be looked at?

Match the cadence to how fast the number can meaningfully move. Weekly for leading indicators, monthly for conversion rates and pipeline, quarterly for anything about efficiency or payback. Looking at a slow number every day produces anxiety and bad decisions.

This is the part that saves teams from themselves. Conversion rates on small numbers swing wildly week to week, so a weekly review invites overreaction. If you are working with tens of conversations rather than thousands, monthly is the honest resolution for most rates.

I would rather have five numbers reviewed on the right cadence than twenty reviewed constantly. Cadence is part of the metric definition, and writing it down next to each tile stops the weekly panic about a monthly number.

Who should the dashboard be built for?

One named person who will act on it. A dashboard built for everyone gets designed for the most senior viewer and used by nobody. If the founder and the person running content need different views, build two small ones rather than one that tries to satisfy both.

The audience decides the framing. A founder wants to know whether the machine is producing pipeline and whether anything is broken. Someone running a channel wants to know whether this week's work moved the thing it was supposed to move. Those are different questions with different tolerances for detail.

This is also why the first report you hand a new client or a new stakeholder matters so much. It sets what they will ask about forever, which is the argument I made in what the first report a new client sees should contain.

What belongs next to the numbers?

A sentence of written commentary, updated on the same cadence. What changed, why you think it changed, and what you are doing about it. Numbers without interpretation get interpreted by whoever is most confident in the room, which is rarely the person closest to the work.

I keep this as a short written note rather than a meeting. Two or three sentences per cycle, stored where the dashboard lives, builds a history you can read later. Six months of those notes is the most useful marketing document most teams never create.

Commentary also protects you from the thing every practitioner fears, which is a number moving for external reasons and being read as your performance. Writing down the context at the time is how you avoid arguing about memory later.

What does a good dashboard look like six months in?

Roughly the same size, with different tiles. If nothing has been removed in six months, nobody is using it critically. A living dashboard sheds numbers that stopped being decisions and gains ones that matter for whatever you are working on now.

I review this at the same time as the quarterly plan, because the leading indicators should follow the current bets. When a channel gets paused, its tile comes off. When a new motion starts, it gets one tile, not five.

The goal is that someone can look at the dashboard for thirty seconds and know whether to worry. That is a high bar, and it is only achievable with a small number of well chosen numbers and someone willing to delete things.

What should you do next?

Open your current dashboard and, for every tile, write down the decision it informs. Delete anything without a decision, add qualified conversations and their source if they are missing, and write one sentence of commentary today.

If you want an outside read on which numbers your team should actually be watching, that is a short conversation with a long payoff. Tell me what you currently report on and I will tell you what I would cut. Let's chat.

Get found, cited and the back office automated

Let's make your site the source AI engines quote and wire up the systems behind it.

Contact

Let's get your website found and cited by AI

Tell me what you're working on, whether AI search is skipping your product, your back office is buried in manual work, or you need a build that does both.

Got it, thanks. I read every message personally and reply within 1-2 business days.
Oops! Something went wrong while submitting the form.